Beat tracking refers to the automated detection of the beat in musical recordings. Two neural network architectures are trained and evaluated for this task.
Descripción:
The beat tracking process is divided into four stages: audio preprocessing, network training, peak-picking and beat detection. In the practical part, two network architectures (CNN and RNN) are trained and evaluated.